Uncertainty in determining ammonia and methane emissions at different sampling locations in an open-sided dairy barn

Abstract

Measuring ammonia (NH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> ) and methane (CH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</inf> ) emissions from open buildings is not straight forward due to high spatial and temporal variabilities of microclimatic variables. This study aims at investigating the uncertainty of emissions at different sampling locations (SLs) in an open-sided dairy barn during spring. Measurements of gas concentrations were continuously acquired by a photo-acoustic analyser. The carbon-dioxide (CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> ) mass balance method was applied to estimate NH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> and CH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</inf> emissions at two groups of SLs (i.e., central and perimeter locations). Different limitation methods based on thresholds for indoor-outdoor CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> differences (i.e., 50 ppm, 70 ppm, 100 ppm) were applied to assess uncertainty in the estimations. Data were analysed through the application of a one-way analysis of variance (ANOVA) for testing significant differences in gas emissions and uncertainties for different groups of SLs. Based on the results, the applied limitation methods modified the data samples available for the analyses with a data destruction of 0.2%, 0.8% and 9.4 % in the central SLs and of 40.7%,63.5% and 86.6% in the perimeter SLs for 50 ppm, 70 ppm and 100 ppm, respectively. Emissions of NH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> and CH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</inf> were significantly different (P<0.001) for the two groups of SLs in all the three limitation methods, when the sample was reduced from 50 ppm to 100 ppm. NH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</inf> emissions were higher in the central SLs than in the perimeter SLs and vice versa for CH <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</inf> emissions. The relative difference of emissions was higher in the perimeter SLs than in the central SLs.